From mboxrd@z Thu Jan 1 00:00:00 1970 From: Boaz Harrosh Subject: Re: [PATCH v2] fix sign extension with 1.5TB usb-storage LBD=y Date: Wed, 22 Apr 2009 10:32:59 +0300 Message-ID: <49EEC82B.5040603@panasas.com> References: <1240347174.10627.20.camel@nimitz> <20090421211858.GA1926@parisc-linux.org> <1240351210.10627.30.camel@nimitz> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1240351210.10627.30.camel@nimitz> Sender: linux-usb-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Dave Hansen Cc: Matthew Wilcox , linux-kernel , mdharm-usb-JGfshJpz5UybPZpvUQj5UqxOck334EZe@public.gmane.org, linux-usb , usb-storage-ijkIwGHArpdIPJnuZ7Njw4oP9KaGy4wf@public.gmane.org, James Bottomley , linux-scsi , viro List-Id: linux-scsi@vger.kernel.org On 04/22/2009 01:00 AM, Dave Hansen wrote: > Here's a patch implementing Al's suggestion. Not quite as trivial as > Matthew's, but even nicer aesthetically. (Description stolen from > Matthew's patch). I have verified that this fixes my issue. > > -- > > Shifting an unsigned char implicitly casts it to a signed int. This > caused 'lba' to sign-extend and Linux would then try READ CAPACITY 16 > which was not supported by at least one drive. Using the > get_unaligned_be*() helpers keeps us from having to worry about how the > extension might occur. > > Signed-off-by: Dave Hansen > > diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c > index 3fcb64b..c50142b 100644 > --- a/drivers/scsi/sd.c > +++ b/drivers/scsi/sd.c > @@ -50,6 +50,7 @@ > #include > #include > #include > +#include > > #include > #include > @@ -1344,12 +1345,8 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, > return -EINVAL; > } > > - sector_size = (buffer[8] << 24) | (buffer[9] << 16) | > - (buffer[10] << 8) | buffer[11]; > - lba = (((u64)buffer[0] << 56) | ((u64)buffer[1] << 48) | > - ((u64)buffer[2] << 40) | ((u64)buffer[3] << 32) | > - ((u64)buffer[4] << 24) | ((u64)buffer[5] << 16) | > - ((u64)buffer[6] << 8) | (u64)buffer[7]); > + sector_size = get_unaligned_be32(&buffer[8]); > + lba = get_unaligned_be64(&buffer[0]); These are actually aligned access it might be worth sacrificing a cast to be32/64 for sake of speed. > > sd_read_protection_type(sdkp, buffer); > > @@ -1400,10 +1397,8 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp, > return -EINVAL; > } > > - sector_size = (buffer[4] << 24) | (buffer[5] << 16) | > - (buffer[6] << 8) | buffer[7]; > - lba = (buffer[0] << 24) | (buffer[1] << 16) | > - (buffer[2] << 8) | buffer[3]; > + sector_size = get_unaligned_be32(&buffer[4]); > + lba = get_unaligned_be32(&buffer[0]); > Here too, both are actually aligned. > if ((sizeof(sdkp->capacity) == 4) && (lba == 0xffffffff)) { > sd_printk(KERN_ERR, sdkp, "Too big for this kernel. Use a " > > > -- Dave > I'm not 100% convinced just a thought. Boaz -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html